Are all Actions Movements of the
Agent's Body?
Julian Fink
Abstract
Davidson famously contended that all actions are movements of
the agent's body. It has been objected, however, that Davidson's
view is incompatible with his own denition of primitive actions.
This paper argues that this objection is based on an incorrect
reading of Davidson's argument.
I will show that by reading
movements, in all actions are bodily movements, transitively,
Davidson's denition of primitive actions ceases to conict with
his thesis that all actions are bodily movements.
Introduction
The present paper champions Davidson's thesis that all actions are movements of the agent's body. I defend this view against Danto's and Feinberg's proposals according to which only basic actions are bodily movements, whereas causally complex actions take place outside the connes
of the agent's body. I argue that Davidson succeeds in showing that all
our actions are basic (or primitive) and that causally complex actions
are identical with basic actions.
However, it has been argued that by endorsing Davidson's denition
of primitive actions one is unable to uphold the thesis that all actions
are bodily movements. For to raise my arm I rst have to bring about
cerebral events and muscle contractions that then cause the rising of my
arm.
Hence, bodily movements are not primitive actions, but are, in
fact, causally complex. So if all our actions are primitive it follows that
none of our actions are movements of the body.
I argue, however, that this conclusion is based on a misreading of
Davidson's argument. The misreading stems from an ambiguity of the
verb to move, when used in a nominal phrase. I argue that if one reads
Kriterion Journal of Philosophy (2011) 24: 5264.
http://www.kriterion.at
c 2011 The author
Julian Fink:
Are all Actions Movements of the Agent's Body?
53
movements, in all actions are bodily movements, transitively, then
Davidson's argument that all actions are primitive is not inconsistent
with the thesis that all our actions are bodily movements. Instead, if all
we do is perform primitive actions, then it follows that all actions are
movements of the agent's body.
1
Suppose I move my ngers, thereby pressing the keys on the keyboard
thereby causing various letters to appear on my screen, which in turn
causes me to perceive what I have just typed.
These are clearly sev-
eral events; the movement of my ngers, the pressing of the keys, the
appearance of the letters on the screen and my perception of these symbols. But even if innitely extended, this chain of causal consequences
would not involve all the events that are causally related to my action.
It would not include, for example, the cerebral events and contractions
of my muscles which take place prior to my nger's movement.
It seems undisputable that most of the events I listed above could not
be described as movements of my body.
Prima facie, only the movement
of my ngers seems to qualify as a bodily movement.
However, we
commonly talk about our actions by making reference to events that are
not bodily movements; I pressed the key, I wrote these words, etc.
If these are genuine actions and they are not somehow identical with
the movements of my nger, then it seems clear that not all actions are
bodily movements.
Two philosophers who deny that my moving my ngers and my pressing the keys are one and the same action are Joel Feinberg and Arthur
Danto.
1 Feinberg proposes that my action of moving my ngers cannot
be identical with my action of typing something that appears on my computer screen. These are not only two (numerically) dierent actions but
also actions of dierent kinds. Whilst my moving my nger is causally
simple, my pressing the keys is a causally complex action. Feinberg
denes a causally simple action as one that is not done by teleologically
2
connected `sub-acts' . In contrast, an action is causally complex if it is
3 For example, [i]n order
done by a teleologically connected `sub-act'.
to open a door, we must rst do something else which will cause the door
to open; but to move one's nger one simply moves it no prior causal
activity is required.
4 Hence, Feinberg proposes that if
complex action and thus eected by another action, say
between
A1
and
A2
is a causal one, i.e.
A1
causes
A2 .5
A2 is a causally
A1 , the relation
Kriterion Journal of Philosophy (2011) 24: 5264
54
Thus, the way Feinberg and Danto construe the relation between
causally simple and causally complex actions entails that not all actions
are bodily movements. If a person opens a door (A2 ) by moving her arm
(A1 ), then
A1
A1
is a distinct action that causes another action
is a bodily movement,
A2
is
not
A2 .
Whilst
a bodily movement. In consequence,
not all actions are movements of the body.
2
One should be suspicious, however, about whether the connection between these two actions ought to be construed as one of cause and eect.
For by supposing that they are related causally, one seems to confuse the
eect
of my action, i.e. the door opening, with the
action
itself, i.e. my
6 Feinberg's example implies that the event that can
opening the door.
be described as my simple action, i.e.
e1 ,
my moving of my arm, say
causes another event that, under the assumed circumstances, can be
described as my action of opening the door, say
e2 .
Let us briey look at the potential and actual causal consequences
e2 may cause a further (complex)
e2 , i.e. the action of opening
the door, will cause an event which is the opening of the door, say e3 .
However, independently of e2 causing e3 , it seems that e1 is causally
sucient to cause e3 . That is, my moving of my arm (e1 ) is the cause of
the door's opening (e3 ). But this means that the event of my moving my
arm (e1 ) not only causes me to open the door (e2 ) which then causes
the door's opening (e3 ) it also immediately causes the door's opening
7
(e3 ).
of
e2 .
According to Feinberg's thesis,
action. Beyond that, it seems also clear that
Unless one believes that one and the same event (e3 ) can be caused
directly by an event (e1 ) whilst at the same time being caused by another event (e2 ) which is itself caused by
e1
one must nd this result
8 Since I think it undeniable that my action of moving my
implausible.
arm (e1 ) causes the opening of the door (e3 ), Feinberg's and Danto's
view that simple acts
cause
causally complex ones cannot be upheld.
Clearly, this does not mean that my action of moving my arm and my
action of opening the door are not somehow related. In the following, I
discuss Davidson's solution to this problem and look at its consequences
for the view that all actions are bodily movements.
Julian Fink:
Are all Actions Movements of the Agent's Body?
55
3
Davidson tackles the problem by arguing that the relation between simple and causally complex actions is one of identity. If I open the door
by moving my arm, my moving of my arm and my opening the door
9 The two verb-phrases moving my arm
are one and the same action.
and opening the door are just varied descriptions of one and the same
descriptum,
namely my action.
10 The dierence between these descrip-
tions lies in the latter's portrayal of the action
qua
a less and a more
remote consequence.
Davidson's argument for this thesis relies on a consideration I discussed above; namely that my moving of my arm is identical with doing
something that causes the door to open. That is, doing something that
causes the door to open is also identical to causing the door to open.
Hence, causing the door to open cannot be something that causes the
causing of the door's opening, as Feinberg suggests. Instead, it
11
is
the
direct cause of the door opening.
Davidson thinks that the urge to separate my action of opening the
door and the action of my moving my arm stems from confusing a feature of the description of an event (i.e. the event being described by
including a reference to its eects) with a feature of the event itself 12 It is often supposed that
(i.e. being the cause of a sequence of events).
describing an event
e2 ,
implies that
e2
e1
by including a reference to one of its consequences
is a part or feature of
e1 .13
This supposition leads to
the questionable result that if my movement of my arm causes the door
to open, the event that is the door's opening is part of the event that
is my movement of my arm. Thus, to avoid this problem, philosophers
like Feinberg and Danto assume that if I open the door by moving my
arm, my arm's movement causes my (action of ) opening the door. Accordingly, my moving my arm then includes the action that causes the
door to open.
However, Davidson shows that this worry is unfounded. If my action
of moving my arm causes the door to close, the event that is the closing of
the door, say
e1 , does not become part of the event that is the movement
e2 . Rather, e1 is a distinct eect of e2 , of which e1 , my
of my body, say
action, is the cause. Thus, my moving my arm is not causing the cause
of the door's opening, but, as Davidson emphasises, it is the cause of the
event that is the door's opening.
Consequently, Davidson identies those actions which Feinberg singles out as causally complex with those that are causally simple. They
are one and the same action. A causally complex action is just a dif-
56
Kriterion Journal of Philosophy (2011) 24: 5264
14
ferent description of an action that is (also) a causally simple action.
But this means that there must be a description for all of our actions
that does not include a reference to their eects. Our actions must be
describable in primitive terms, i.e., without including events that are
15 Davidson holds that these primitive descrip-
caused by our actions.
tions of our actions are always constituted of bodily movements. This is
because, when I move my nger, I do not do so by doing something that
causes my moving of my nger; rather [d]oing something that causes my
nger to move [. . . ]
is
16 Thus Davidson holds that
moving my nger.
our primitive actions, the ones we do by not doing something else, [are]
17 That is to say, whenever we describe
mere movements of the body.
an action without including a reference to its eects, what we end up
with is a description of a bodily movement.
this that, in fact,
all
we
do
Davidson concludes from
is to move our bodies.
He expresses this
view by supposing that if the queen kills the king by moving her hand in
such a way that she pours poison in his ear, it would be absurd to ask,
after the queen has moved her hand in such a way as to cause the king's
death, [whether] any deed remains for her to be done or to complete.
For [s]he has done her work; it only remains for the poison to do its.
18
Hence, our primitive actions are all the actions there are. We never do
more than move our bodies the rest is up to nature.
19
In sum, Davidson argues that: (i) all primitive actions are bodily
movements; and (ii) all our actions are primitive. By combining (i) and
(ii) it then follows that all actions are movements of the body.
4
Some philosophers argue that the conclusion that all actions are movements of the body is incompatible with one of Davidson's arguments
concerning the intentionality of internal events that are necessary to
20
cause bodily movements.
A man who raises his arm both intends to do so with his
body whatever is needed for his arm to go up and knows
that he is doing so. And of course the cerebral events and
movements of his muscles are just what is needed. So, though
the agent may not know the names or locations of the relevant muscles, nor even that he has a brain, what he makes
happen in his brain and muscles when he moves his arm
is, under one natural description, something he intends and
knows about.
21
Julian Fink:
Are all Actions Movements of the Agent's Body?
57
That is, if I raise my arm, this action includes my intentional action of
bringing about the event that causes my arm to rise. James Montmarquet points out that, considered on its own, Davidson's remark could
22 However, since Davidson
have at least three distinct interpretations.
identies doing something that causes my nger to move with moving
my nger, what he must intend to say here is that my raising my arm is
identical with doing something (or bringing about an event) that causes
23
my arm to rise.
Montmarquet argues that this view is problematic. For it implies that
arm-raisings are identical with acts which cause arm risings.
24 That is
to say, the event that causes my arm to rise is identical with its eect,
i.e. the event of the actual arm rising. However, since cause and eect
cannot be one and the same event, Davidson's contention that doing
something that causes my nger to move
is
moving my nger turns out
to be untenable. In fact, Davidson himself, in another example, seems
to admit implicitly that this is so. He argues that to trip my self [. . . ]
is not identical with what it causes [i.e. my tripping].
25 So what causes
my tripping and my tripping must be two dierent events.
If this is
correct, the same must also hold for causing a bodily movement and the
bodily movement itself. But if the event of an agent's bodily movement
is distinct from its cause and is thus not included in the event that is
the movement of the arm, it seems clear that the cause of the arm's
movement must; (i) be prior to bodily movement; and (ii) be located
inside the body of the agent.
Montmarquet, amongst others
26 , argues that the non-identity be-
tween the cause of a bodily movement and the bodily movement itself
implies that Davidson cannot maintain that bodily movements are primitive actions. For when I move my arm, I do not do so without doing
something else rst. That is to say, to raise my arm, I must bring about
internal events (neural events and muscle contractions) that then cause
my arm to rise.
Thus, the raising of my arm is done by my bringing
about these internal causes. According to the Davidson-Anscombe thesis
27 , bringing about these internal events and my moving my arm must
be regard as one and the same action; saying I moved my body simply
describes my action by including a reference to one of its consequences.
But this means that my bodily movements must be typied as causally
complex and not as primitive actions.
Thus Montmarquet concludes
that [a]cts of bringing about bodily movements are not identical with
the bodily movements in whose bringing about they consist.
28 The act
of my raising my arm is thus an internal event that causes my arm to
Kriterion Journal of Philosophy (2011) 24: 5264
58
rise, but not the event that is the rising of my arm itself.
Consequently, bodily movements are not primitive actions. Combining this with the second of Davidson's conclusions that I discussed above,
i.e. all actions are primitive, implies that no action is a movement of the
agent's body. The picture we get can be described as follows.
When I move my hand, the movement of my hand, though
an eect of my action, is not itself an action, and no one
who considered the matter would say it was any more than
he would say that the death of Caesar, as distinct from his
murder, was an action or even part of an action.
29
Accordingly, the action itself is not identical with the movement of my
arm, but with the event that causes the movement of my arm.
However, even though this picture may seem convincing, I do not
think that it is an inevitable consequence of Davidson's view that none
of our actions are bodily movements.
In the following, I argue that
Davidson's argument only leads to this result if it receives a profound
misreading.
5
Davidson argues that when I move my hand, thereby (say) pushing a key
which causes a letter to appear on my computer screen, I have performed
one
action that can be variously described by including references to its
more or less remote consequences. However, this is not to say that my
action is either identical with the events that are the appearance of the
letter or the key being pressed. Instead, Davidson identies the action
with the event that is my movement of the nger. He expresses this view
by saying that [d]oing something that causes my nger to move [. . . ] is
moving my nger. The problem of this view is that it seems to identify
the event that is the movement of my nger with its cause. However,
since cause and eect can not be one and the same event, my doing
something that causes my body to move cannot be a bodily movement
itself.
But this implies, as I discussed above, that bodily movements
cannot be primitive actions; yet since, according to Davidson, all we do
is perform primitive actions, bodily movements cannot be actions at all;
or so it is argued.
But I think that, despite the last quote, Davidson is not committed to
the view that an action can be identied with the event that is the bodily
movement, at least in the way that leads to the alleged cause/eect
Julian Fink:
identity.
Are all Actions Movements of the Agent's Body?
59
This interpretation is based on a misreading of Davidson's
30
argument that stems from an ambiguity of the verb to move.
The verb to move (cognate verbs, such as to rise, to boost, to
melt, etc.) is ambiguous between its transitive and intransitive forms.
In I moved my body, move occurs transitively; in my body moved,
move occurs intransitively. Embedded in a whole sentence, its usage is
normally revealed through the context; yet when it occurs in a nominal
phrase, it proves ambiguous.
31 That is, moving my nger may signify
a transitive relation between me and my nger's movement, indicating
that I cause my nger to move. Or it may signify no relation, in which
case it refers to the event that is the motion of my nger. (To avoid any
further ambiguity, I shall follow Hornsby by using the subscripts T and
32 )
I to distinguish the transitive and the intransitive forms.
The philosophical signicance of this ambiguity springs from the connection that holds between what is expressed by the transitive form and
the intransitive forms. For if a billiard ball (B1 ) movesT another billiard
ball (B2 ), this implies that
B1
causes
B2
33 Clearly, the same
to moveI .
relation holds between I moveT my body and my body movesI . Thus
I moveT my body implies that I cause my body to moveI . By assuming
that agent causation is reducible to event causation, saying I movedT
my arm describes the event that causes my arm to move by including
a reference to (one of its) eects (i.e.
the movementI of my body).
34
That is, I moveT my arm is a description of the event that causes my
arm to moveI . But does it then follow that Davidson is committed to
the proposition that the action that is the movingT of my arm is identical with the event of my arm movingI , as it is assumed by Davidson's
critics?
I think the answer to this question is no if one interprets the rst
occurrence of the verb to move in [d]oing something that causes my
nger to move [. . . ]
is
moving my nger as being used transitively and
the second move as being used intransitively. For if I startle my friend
by opening the door, though the opening of the door and the startling of
my friend are one and the same action, Davidson's view does not imply
that the event that is the startling of my friend is identical with the
event that is the opening of the door. Instead, these are separate events
that are causally connected and caused by my action. The same holds
for the case where I moveI my arm by movingT my arm. The event of
my movingT my arm is not identical with the event of my arm movingI .
Recall, however, that Davidson's critics premised their verdict that for
Davidson
no
action can be bodily on Davidson's view that moving my
Kriterion Journal of Philosophy (2011) 24: 5264
60
armT is identical with my arm movingI . This is precisely what was
supposed to follow from Davidson's remark that [d]oing something that
causes my nger to move [. . . ] is moving my nger. However, as I stated
above, Davidson holds that doing something that causes my nger to
move is identical with causing my nger to move. Combining this with
what we have just established, namely that I moveT my arm refers to
the cause of my arm's movingI , all that Davidson seems to suggest here is
that the movingT of my nger is the movingT of my nger something
that is hardly deniable and apparently does not identify the cause of
a bodily movementI with the bodily movementI itself.
Accordingly,
Davidson's view does not lead to any contradictory results.
Admittedly, this result depends on a particular interpretation of
Davidson's proposition that [d]oing something that causes my nger
to move [. . . ]
is moving my nger.
This statement only leads to a
non-contradictory result if the rst moving in this statement is used
intransitively and the second is used transitively.
But why should we
think that this is the way Davidson intended to use the verb to move
in this sentence?
I think Davidson answers this question in his Ac-
tions, Reasons, and Causes where he writes: Obviously, the problem
is greatly aggravated if we assume, as Melden does [. . . ] that an action
(raising one's arm) can be identical with a bodily movement (one's arm
going up).
35 To avoid this mistake himself, Davidson
say that the something I do that causes my nger to
ing T
must intend to
move I is my mov-
my nger. In consequence, Davidson's statement does indeed not
lead to a contradictory result.
In sum, neither Davidson's statement that doing something that
causes my nger to move is moving my nger, nor his general argument
should be taken to imply that the cause of my bodily movementI is
identical with its eect, i.e.
the movementI of the body.
It should
rather be understood as the claim that my doing something that causes
the movementI of my nger is my movingT my nger, i.e. bringing about
internal events that cause my nger to moveI .
6
This paper championed the thesis that all actions are movementsT of the
agent's body. I defended this thesis against the view that our primitive
actions, which are bodily movements, cause complex actions, which are
not bodily movements. Following Davidson, I argued that my startling
my friend by opening the door, does not entail that my action of opening
Julian Fink:
Are all Actions Movements of the Agent's Body?
61
the door causes my action of startling my friend. Instead of causing a
connected sequence of acts, my opening of the door causes a sequence
of events, some of which can be picked out to describe my action. Considering all possible descriptions of an action, some descriptions will be
primitive in the sense that they refer to it without including any of
its consequences. These primitive actions can be identied with bodily
movements, because doing something that causes the movement of my
nger
is
just moving my nger. Furthermore, since every action has a
description under which it is a primitive action, it follows that all actions
are movements of the body.
It has been argued, however, that by embracing Davidson's view,
bodily movements cannot be described as primitive actions. This is so
because bodily movements do not t Davidson's denition of a primitive
action. For if I move my nger I must do something that causes my nger
to move, namely contract my muscles. Further, to contract my muscles I
must do something which causes my muscles to contract, etc. However,
making these events occur is clearly not any bodily movement.
Thus,
bodily movements are (only) the eect of my actions so they cannot be
classied as primitive actions. But since, according to Davidson, all we
do is perform primitive actions, it follows that no action is a movement
of an agent's body.
However, I argued that a sound interpretation of Davidson's argument does not lead to this conclusion. This interpretation is based on
clarifying an ambiguity of the verb to move, which essentially allows
one to say that to moveT my nger refers to an event which causes my
nger to moveI . Thus, if internal brain events and muscle contractions
are causing my nger to moveI , then making these internal events occur
is movingT my nger. Accordingly, the fact that my action causes the
movementI of my nger does
not
imply that the movementT of my n-
ger is caused by one of my actions. Instead, my movingT my nger is a
primitive action of mine. In combination with Davidson's thesis that all
actions are primitive, this leads to an unequivocal upshot:
are movementsT of the agent's body.
all
actions
Kriterion Journal of Philosophy (2011) 24: 5264
62
Notes
1 [8]; [4]
2 [8, p.145]
3 Danto draws a similar distinction, by distinguishing between basic actions and
those which are caused by basic actions [4].
4 [8, p.147]; Feinberg's emphasis).
5 [8]; Davidson thinks that [3] commits to the same mistake.
6 Cf. [5, p.56].
7 By saying that
e1
causes
e3
directly, I mean that
causing another event rst, which then causes
e1
causes
e3
not in virtue of
e3 .
8 In fact, for anyone who believes that all events have their properties
(i.e.
necessarily,
e1
and
e2
essentially,
are dierent, if they have a dierent cause [cf.
[5,
p.179]; [7, p.17]]), this result must appear as a clear contradiction.
9 This thesis is sometimes referred to as the Davidson-Anscombe thesis. Anscombe
introduced this view in her book
Intention
(see [2, pp.37-47].
10 Cf. [5, p.59].
11 Cf. [5, p.58]
12 [5, p.58]
13 [5, p.58]
14 One exception to this view is the case where an agent, for example, moves her
right arm by moving her left arm. For then the causally complex action seems
to occur within the connes of the body.
15 Cf. [5, p.49]
16 [5, pp.49f]. Also, cf. [6, p.37] and [14, p.199].
17 [5, p.59].
18 [5, p.58]
19 [5, p.59]
20 See [10] and [1].
21 [5, p.50]
22 [10, p.137]
23 Section V argues that this is, in fact, a misinterpretation of Davidson's argument.
24 [10, p.138]
25 [5, p.47]
26 Cf. [1, p.229].
27 See note 9.
28 [10, p.140]
29 [12, p.191]
30 [9]. See also [11].
31 [9, p.2]
32 Cf. [9, p.2].
Julian Fink:
Are all Actions Movements of the Agent's Body?
63
33 Cf. [1, p.223, n.9].
34 Cf. [11, p.33]. For a critique of the assumption that agent causation is reducible
to event causation see [13].
35 [5, p.5, n.2]
Julian Fink
Universität Wien
Institut für Philosophie Forschungsbereich 2
Dr.-Karl-Lueger-Ring
1010 Wien, Austria
<[email protected]>
64
Kriterion Journal of Philosophy (2011) 24: 5264
References
[1] Maria Alvarez and John Hyman. Agents and their actions.
ophy,
Philos-
73:pp.219245, 1998.
Intention.
[2] G.E.M. Anscombe.
Basil Blackwell, Oxford, 1957.
[3] John Langshaw Austin. A plea for excuses. In
Philosophical Papers,
pages 102122. Clarendon Press, Oxford, 1961.
[4] Arthur Danto.
Basic actions.
American Philosophical Quarterly,
2:pp.141148, 1965.
Essays on Actions and Events.
[5] Donald Davidson.
Clarendon Press,
Oxford, 1980.
[6] Donald Davidson. Problems in the explanation of actions. In R. Sylvan Pettit and J. Norman, editors,
Metaphysics and Morality. Basil
Blackwell, Oxford, 1987.
[7] Donald Davidson. Thinking causes. In J. Heil and A. Mele, editors,
Mental Causation,
[8] Joel Feinberg.
pages 317. Clarendon Press, Oxford, 1993.
Action and responsibility.
D. D. Merrill, editors,
In W. H. Capitan and
Philosophy in America,
pages 134160. Uni-
versity of Pittsburgh Press, Pittsburgh, 1965.
[9] Jennifer Hornsby.
Actions.
[10] James Montmarquet.
Routledge & Kegan Paul, London, 1980.
Actions and bodily movements.
Analysis,
38:pp.13740, 1978.
[11] D.W.D. Owen.
Analysis,
Actions and bodily movements:
Another move.
40:pp.3335, 1980.
[12] H. A. Prichard.
Moral Obligation.
Oxford University Press, Oxford,
1968.
[13] Richard Taylor.
Action and Purpose.
Prentice Hall, Englewood
Clis, 1966.
[14] Irving Thalberg. Do we cause our own actions?
201, 1967.
Analysis, 27:pp.196